- 26 Nov, 2008 37 commits
-
-
Alexey Dobriyan authored
State GC is per-netns, and this is part of it. Signed-off-by:
Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Alexey Dobriyan authored
km_waitq is going to be made per-netns to disallow spurious wakeups in __xfrm_lookup(). To not wakeup after every garbage-collected xfrm_state (which potentially can be from different netns) make state GC list per-netns. Signed-off-by:
-
Alexey Dobriyan authored
All of this is implicit passing which netns's hashes should be resized. Signed-off-by:
-
Alexey Dobriyan authored
Signed-off-by:
-
Alexey Dobriyan authored
Since hashtables are per-netns, they can be independently resized. Signed-off-by:
-
Alexey Dobriyan authored
Signed-off-by:
-
Alexey Dobriyan authored
Signed-off-by:
-
Alexey Dobriyan authored
Signed-off-by:
-
Alexey Dobriyan authored
This is done to get a) simple "something leaked" check b) cover possible DoSes when other netns puts many, many xfrm_states onto a list. c) not miss "alien xfrm_state" check in some of list iterators in future. Signed-off-by:
-
Alexey Dobriyan authored
To avoid unnecessary complications with passing netns around. * set once, very early after allocating * once set, never changes For a while create every xfrm_state in init_net. Signed-off-by:
-
Alexey Dobriyan authored
Signed-off-by:
-
Alexey Dobriyan authored
Signed-off-by:
-
Ingo Molnar authored
fix warning: drivers/net/3c523.c:582: warning: ‘cleanup_card’ defined but not used No code changed: md5: ebe4a1b27d3f21b0b12a78c58463b0d7 3c523.o.before.asm ebe4a1b27d3f21b0b12a78c58463b0d7 3c523.o.after.asm Signed-off-by:
Ingo Molnar <mingo@elte.hu> Signed-off-by:
-
Ingo Molnar authored
fix warning: drivers/net/depca.c: In function ‘depca_eisa_probe’: drivers/net/depca.c:1564: warning: ‘mem_start’ may be used uninitialized in this function this seems to be a real bug - depca_eisa_probe() does not check for failure. Add it, symmetric to depca_isa_probe(). Signed-off-by:
-
Ingo Molnar authored
fix this warning: drivers/net/atlx/atl2.c: In function ‘atl2_request_irq’: drivers/net/atlx/atl2.c:644: warning: unused variable ‘err’ 'err' is unused in the !CONFIG_PCI_MSI case. Instead of further increasing the #ifdeffery in this function, restructure the code a bit and get rid of the #ifdef. This relies on the fact that pci_enable_msi() will always fail in the !CONFIG_PCI_MSI case. There should be no change in driver behavior. Signed-off-by:
-
Ingo Molnar authored
fix this warning: net/bluetooth/rfcomm/sock.c: In function ‘rfcomm_sock_ioctl’: net/bluetooth/rfcomm/sock.c:795: warning: unused variable ‘sk’ perhaps BT_DEBUG() should be improved to do printf format checking instead of the #ifdef, but that looks quite intrusive: each bluetooth .c file undefines the macro. Signed-off-by:
-
Ingo Molnar authored
fix this warning: net/sunrpc/xprtrdma/verbs.c: In function ‘rpcrdma_conn_upcall’: net/sunrpc/xprtrdma/verbs.c:279: warning: unused variable ‘addr’ Signed-off-by:
-
Ingo Molnar authored
fix this warning: net/ax25/sysctl_net_ax25.c:27: warning: ‘min_ds_timeout’ defined but not used net/ax25/sysctl_net_ax25.c:27: warning: ‘max_ds_timeout’ defined but not used These are only used in the CONFIG_AX25_DAMA_SLAVE case. Signed-off-by:
-
Ingo Molnar authored
fix warning: drivers/net/mlx4/mcg.c: In function ‘mlx4_multicast_attach’: drivers/net/mlx4/mcg.c:217: warning: integer overflow in expression there was no real danger of overflow here though. md5: db8eb55620f886c03854a2abb2ce6c3f mcg.o.before.asm db8eb55620f886c03854a2abb2ce6c3f mcg.o.after.asm Signed-off-by:
-
Ingo Molnar authored
this warning: net/dccp/options.c: In function ‘dccp_parse_options’: net/dccp/options.c:67: warning: ‘value’ may be used uninitialized in this function is a bogus GCC warning. The compiler does not recognize the relation between "value" and "mandatory" variables: the code flow can ever reach the "out_invalid_option:" label if 'mandatory' is set to 1, and when 'mandatory' is non-zero, we'll always have 'value' initialized. Help out the compiler by annotating the variable. Signed-off-by:
-
Ingo Molnar authored
this warning: drivers/net/cassini.c: In function ‘cas_rx_ringN’: drivers/net/cassini.c:2350: warning: ‘skb’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between cas_rx_process_pkt() and 'skb'. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: drivers/net/mlx4/en_rx.c: In function ‘mlx4_en_activate_rx_rings’: drivers/net/mlx4/en_rx.c:412: warning: ‘err’ may be used uninitialized in this function Triggers because 'err' is uninitialized in the following input conditions: priv->rx_ring_num is zero and mlx4_en_fill_rx_buffers() fails. But even if ->rx_ring_num is nonzero, 'err' will be zero if mlx4_en_fill_rx_buffers() fails and mlx4_en_activate_rx_rings() returns success - incorrectly. So it's best to keep the error code uptodate on mlx4_en_fill_rx_buffers() calls as well. Signed-off-by:
-
Ingo Molnar authored
this warning: drivers/net/wan/z85230.c: In function ‘z8530_interrupt’: drivers/net/wan/z85230.c:713: warning: ‘intr’ may be used uninitialized in this function is clearly bogus - annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: drivers/net/sis900.c: In function ‘sis900_timer’: drivers/net/sis900.c:1280: warning: ‘speed’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between sis900_read_mode(), 'speed' and 'duplex'. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: fs/dlm/netlink.c: In function ‘dlm_timeout_warn’: fs/dlm/netlink.c:131: warning: ‘send_skb’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between prepare_data() and send_skb. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: net/dsa/mv88e6060.c: In function ‘mv88e6060_poll_link’: net/dsa/mv88e6060.c:225: warning: ‘port_status’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between 'link' and 'port_status'. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: net/dsa/mv88e6xxx.c: In function ‘mv88e6xxx_poll_link’: net/dsa/mv88e6xxx.c:361: warning: ‘port_status’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between 'link' and 'port_status'. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: net/ipv6/ip6_flowlabel.c: In function ‘ipv6_flowlabel_opt’: net/ipv6/ip6_flowlabel.c:467: warning: ‘err’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between fl_create() and 'err'. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: net/sched/sch_hfsc.c: In function ‘hfsc_enqueue’: net/sched/sch_hfsc.c:1577: warning: ‘err’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between hfsc_classify(), 'cl' and 'err'. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: net/sunrpc/xprtrdma/svc_rdma_transport.c: In function ‘svc_rdma_accept’: net/sunrpc/xprtrdma/svc_rdma_transport.c:830: warning: ‘dma_mr_acc’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) flow connection between need_dma_mr and dma_mr_acc. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: drivers/net/qla3xxx.c: In function ‘ql3xxx_probe’: drivers/net/qla3xxx.c:3912: warning: ‘pci_using_dac’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between 'pci_using_dac' and 'err'. Annotate it. Signed-off-by:
-
Ingo Molnar authored
this warning: drivers/net/niu.c: In function ‘esr_reset’: drivers/net/niu.c:741: warning: ‘reset’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between: - esr_read_reset() and 'reset' Annotate it. Signed-off-by:
-
Ingo Molnar authored
these warnings: drivers/net/niu.c: In function ‘serdes_init_niu_1g_serdes’: drivers/net/niu.c:451: warning: ‘sig’ may be used uninitialized in this function drivers/net/niu.c: In function ‘serdes_init_niu_10g_serdes’: drivers/net/niu.c:550: warning: ‘sig’ may be used uninitialized in this function triggers because GCC does not recognize that the max_retry loop always initializes 'sig', due to max_retry != 0. Annotate them. Signed-off-by:
-
Ingo Molnar authored
this warning: drivers/net/s2io.c: In function ‘rx_intr_handler’: drivers/net/s2io.c:7369: warning: ‘lro’ may be used uninitialized in this function triggers because GCC does not recognize the (correct) error flow between: - s2io_club_tcp_session()and 'lro' Annotate it. Signed-off-by:
-
Daniel Lezcano authored
This patch will filter out the uevent not related to the init_net. Without this patch if a network device is created in a network namespace with the same name as one network device belonging to the initial network namespace (eg. eth0), when the network namespace will die and the network device will be destroyed, an event will be sent and catched by the udevd daemon. That will result to have the real network device to be shutdown because the udevd/uevent are not namespace aware. Signed-off-by:
Daniel Lezcano <dlezcano@fr.ibm.com> Acked-by:
"Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by:
-
Robert Olsson authored
Signed-off-by:
Robert Olsson <robert.olsson@its.uu.se> Acked-by:
Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by:
-
Robert Olsson authored
A patch to put names on the niu interrupt vectors according the syntax below. This is needed to assign correct affinity. > So on a multiqueue card with 2 RX queues and 2 TX queues we'd > have names like: > > eth0-rx-0 > eth0-rx-1 > eth0-tx-0 > eth0-tx-1 Signed-off-by:
Jesper Dangaard Brouer <jdb@comx.dk> Signed-off-by:
-
- 25 Nov, 2008 3 commits
-
-
Ilpo Järvinen authored
Since pskb_expand_head creates copy of the shared area we cannot keep any frag ptr past de-cloning. This fixes the tcpdump recvfrom -EFAULT problem. Signed-off-by:
Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by:
-
Jarek Poplawski authored
After implementing qdisc->ops->peek() there is no more calling qdisc_tree_decrease_qlen() without rtnl_lock(), so qdisc_list_lock added by commit: f6e0b239 "pkt_sched: Fix qdisc list locking" can be removed. Signed-off-by:
Jarek Poplawski <jarkao2@gmail.com> Signed-off-by:
-
Eric Dumazet authored
Impact: Optimization Like done in inet_unhash(), we can avoid taking a chain lock if socket is not hashed in udp_unhash() Triggered by close(socket(AF_INET, SOCK_DGRAM, 0)); Signed-off-by:
Eric Dumazet <dada1@cosmosbay.com> Signed-off-by:
-
